Naegele's Rule and Why Your Due Date Is an Estimate, Not a Deadline

A Formula From 1812 Is Still Telling You When Your Baby Will Arrive

Every pregnancy due date in the modern world starts with the same arithmetic: take the first day of the last menstrual period, add a year, subtract three months, add seven days. That is Naegele's rule. It is named after Franz Karl Naegele, a German obstetrician who wrote it down in a textbook in 1812. The math has not been updated in two hundred and fourteen years, and the assumptions baked into it would not survive a single afternoon of honest scrutiny by anyone who has ever tracked their own cycle.

None of this is a scandal. Naegele's rule is a useful approximation, and the medical system has learned to wrap it in caveats. What is more interesting is how few of those caveats actually reach the people whose pregnancies are being dated. A due date arrives on the page as a single, specific calendar day, and that specificity does work on the human brain that the underlying statistics do not support. People plan around it. They feel late on the day after it. Employers schedule around it. And only about one in twenty babies actually arrives on it.

This piece is the honest version of the conversation. Where the rule comes from, what it assumes, what the real distribution of spontaneous labor looks like, and how to read an estimated due date as the rough probability cloud it is, rather than the deadline it pretends to be.

Who Was Naegele, and What Did He Actually Write?

Franz Karl Naegele was born in 1778 and ran the obstetrics department at the University of Heidelberg from 1810 until his death in 1851. He wrote a widely used textbook, Lehrbuch der Geburtshilfe, in which he described what would later carry his name as a method for predicting the date of confinement, or Geburtstermin.

His formula was: add seven days to the first day of the last menstrual period, then count back three months. In practice that is identical to adding 280 days, or forty weeks. He attributed the underlying observation to the Dutch physician Herman Boerhaave, who had described a similar method in the early 1700s. Naegele was crediting an even older tradition; the historical record suggests the forty-weeks-from-the-period idea was already informal practice among European midwives well before either man wrote it down.

What is worth pausing on is what Naegele was actually trying to do. He needed a quick way to give a woman a date she could plan around. He worked in an era with no ultrasound, no reliable ovulation detection, no understanding of luteinizing-hormone surges, and no statistical apparatus for thinking about distributions. He needed a single number, and a single number is what he produced. The medical world adopted it because it was simple, it was teachable, and it was approximately correct for the population that Naegele saw.

The Assumptions Hiding Inside the 280 Days

Naegele's rule encodes several specific assumptions about the human body. They are not bad assumptions for a 1812 average, but they are not true for every body, and in some cases they are not even true on average.

The first assumption is that the menstrual cycle is exactly twenty-eight days long. This number is the source of the seven-day addition: if ovulation happens on day fourteen of a twenty-eight-day cycle, and pregnancy lasts 266 days from conception, then 266 plus 14 gives 280 days from the first day of the last period. Change the cycle length and the arithmetic shifts. A person with a regular thirty-five-day cycle ovulates closer to day twenty-one, which means the formula understates their pregnancy by a full week if you do not adjust for it. Most modern calculators, including the one we ship at our pregnancy due date calculator, ask for cycle length and add the difference from twenty-eight to the due date. That is a small correction that gets the answer noticeably closer for anyone whose cycle is not the textbook length.

The second assumption is that ovulation happens reliably on day fourteen of any cycle. It does not. The luteal phase, which is the second half of the cycle, is fairly constant at around fourteen days in most people, but the follicular phase is highly variable. Stress, illness, travel, breastfeeding, weight changes, and ordinary biological variation can shift the ovulation day by a week or more without the cycle length looking unusual at all. A formula that derives the date from the period rather than the conception is, in effect, dating the pregnancy from an event that happened at a variable distance from the one that actually mattered.

The third assumption is that 280 days is the right average pregnancy length in the first place. This one has been studied carefully, and the answer is "mostly, but not exactly." A 1990 study by Mittendorf and colleagues in Obstetrics & Gynecology looked at white nulliparous women (first-time mothers) with reliable cycle data and found the average spontaneous-labor date was closer to 41 weeks and 1 day, not 40 weeks. Subsequent studies have found smaller deviations in other populations, but the consistent finding is that Naegele's 280-day number is a slight underestimate of average gestational length, and the underestimate is larger for first pregnancies than for subsequent ones.

The fourth assumption is that the date of the last menstrual period is precisely known. In practice, plenty of people are not certain to the day. Some have irregular cycles. Some have spotting between periods. Some conceive while breastfeeding or coming off hormonal contraception, with no clean reference point at all. A formula that takes the LMP as a fixed input is implicitly assuming a level of self-knowledge that not everyone has.

What the Actual Distribution Looks Like

The more honest way to think about a due date is as the center of a probability distribution, not as a target. Spontaneous labor onset in healthy pregnancies follows a roughly bell-shaped curve centered slightly after forty weeks, with most births falling in a two-week window on either side. About one in twenty babies arrives on the exact estimated due date. Roughly two-thirds arrive within a week of it. Around eighty percent arrive within two weeks either side. That is the rough shape of what you are actually looking at when you see "October 14" on a chart.

The American College of Obstetricians and Gynecologists formalized this in a 2013 redefinition, reaffirmed in 2017. They drew lines that the older language did not: a baby born from 37 weeks 0 days to 38 weeks 6 days is "early term." From 39 weeks 0 days to 40 weeks 6 days is "full term." From 41 weeks 0 days to 41 weeks 6 days is "late term." From 42 weeks onward is "post-term." This was a deliberate move away from treating the whole 37-to-42-week range as interchangeably "term," because the outcomes are not interchangeable. The point for the present discussion is that the modern clinical vocabulary already acknowledges that delivery is a range, not a date.

Postdates pregnancy is a separate clinical question. Once you pass 41 weeks, the obstetric calculus changes and induction is often discussed. Past 42 weeks, the risk profile shifts further. None of that is contradicted by the observation that the due date is an estimate; it just means that "estimate" has a boundary, on the late side, where it becomes a clinical decision rather than a calendar one.

Ultrasound Dating and Why It Often Wins

The modern correction to Naegele's rule is first-trimester ultrasound. A crown-rump-length measurement between roughly seven and thirteen weeks gives a gestational age accurate to within about five to seven days. After fourteen weeks the accuracy widens. By the third trimester, ultrasound dating can be off by three weeks or more, because babies grow at different rates and the measurement is no longer reading a tight developmental clock.

ACOG's guidance, in its 2017 Committee Opinion 700, is that the LMP-based date and the ultrasound-based date should be compared, and the ultrasound date should be used if it differs from the LMP date by more than a defined threshold. The threshold tightens as the ultrasound is done later. Early ultrasound trumps a sketchy LMP; a confident LMP can sometimes hold up against a late ultrasound. The point is that the date you receive at a prenatal appointment is usually a synthesis, not a pure application of Naegele's rule, and the synthesis is more accurate than either input alone.

This is also why a tool that asks only for LMP and cycle length is giving you a useful estimate but not a clinical one. It cannot incorporate the ultrasound information your provider has. It is closer to a back-of-envelope number than a medical record. If you have had a dating ultrasound and the result differs from what an LMP-only calculator gives you, the ultrasound is almost certainly closer to the truth.

Why the Single-Date Framing Persists Anyway

If a due date is really a probability cloud, why does the system keep giving people a single calendar day? Three reasons, all of them practical rather than scientific.

The first is that humans plan in dates, not in distributions. Maternity leave, employer notification, childcare arrangements, family travel, and a hundred other logistics need a focal point. "Sometime in mid-October, give or take two weeks" does not slot into a calendar app. The single date is a coordination device.

The second is that obstetric scheduling is built around gestational age. Every prenatal appointment, screening test, and clinical decision is timed against weeks-of-gestation, and weeks-of-gestation is computed forward from the estimated due date. The system needs a stable anchor, and a single date provides one. If providers had to operate on a distribution at every appointment, the schedule would not function.

The third is that the alternative is harder to communicate. A 5%-chance-on-the-due-date framing is statistically correct and emotionally exhausting. The single date, with quiet caveats from a provider who has done this a thousand times, is what most patients want. The trade-off is that some patients will internalize the date as a deadline and feel a small failure of biology when they pass it. That is not a problem the medical system has fully solved.

How to Read a Due Date Without Being Misled By It

None of this is reason to abandon due dates. They are useful, and they are roughly right. The honest reframe is to treat the due date as the peak of a probability distribution that you should imagine as a roughly four-week window centered slightly after forty weeks from your last period. Within that window, the baby will arrive at some point that no formula and no ultrasound can pin down in advance. The job of an estimate is to focus planning, not to fix the future.

The practical version of this for anyone using a back-of-envelope calculator: get the cycle-length adjustment right, treat the number as a midpoint rather than a target, do not compare yourself to the date if you pass it by a few days, and use the actual clinical date from your provider once you have it, because they will have folded in ultrasound dating you cannot replicate at home. A 1812 formula running on a 2026 calendar is going to be approximately right. Approximately is enough to plan around. It is not enough to feel late about.